No Two Alike

by Judith Rich Harris

  Forming stereotypes, learning categories, and figuring out the rules of grammar all depend on storing up and putting together data from multiple observations—something the mind does automatically, without any orders from the captain. But once it’s there, the rule or stereotype can often be fished up by the conscious mind. Even if you never thought about it before, you could probably tell me the rule for forming the past tense of a regular verb. Novelists and playwrights make use of the availability of social stereotypes, especially for their minor characters, and so did Sherlock Holmes. The great detective was a master of disguise but he didn’t disguise himself as a particular person: he disguised himself as a particular type of person—a stereotype. Here’s a passage from a story called The Sign of Four, narrated by the faithful Dr. Watson:

  A heavy step was heard ascending the stair, with a great wheezing and rattling as from a man who was sorely put to it for breath. Once or twice he stopped, as though the climb were too much for him, but at last he made his way to our door and entered. His appearance corresponded to the sounds which we had heard. He was an aged man, clad in seafaring garb, with an old pea-jacket buttoned up to his throat. His back was bowed, his knees were shaky, and his breathing was painfully asthmatic. As he leaned upon a thick oaken cudgel his shoulders heaved in the effort to draw the air into his lungs. He had a colored scarf round his chin, and I could see little of his face save a pair of keen dark eyes, overhung by bushy white brows and long gray side-whiskers.17

  Watson was fooled: it was Sherlock Holmes himself. The disguise was aimed squarely at the social category “aged man,” with a few extra touches to provide the seafaring theme. The ancient mariner—minus, of course, the albatross.

  The construction of stereotypes begins early. The mommies and daddies young children pretend to be in their imaginative games are stereotypical adults, not real ones. Their real mommies and daddies are sometimes perplexed when, for example, a four-year-old girl whose own mother is a physician insists that girls can become nurses but only boys can become doctors.18 This anecdote was reported some years ago, at a time when female physicians were still uncommon. The little girl is a grownup now and no doubt her stereotypes have changed. But no doubt she still has stereotypes.

  Research on adults has shown that people who deny being prejudiced may have negative stereotypes of other races. However, this research almost always involves showing subjects pictures (or giving them ethnically identifiable names) of strangers. If you have no other information about a person, the categorization mechanism kicks in, and it does its work largely on an unconscious level.19 But once you have established a page in your people-information lexicon for a particular individual, the relationship system clamors to be heard. Now you have two different systems telling you things about the same individual.

  One of the clues that two different systems are involved in regulating behavior is that they occasionally issue conflicting orders. The conflicts between the categorization mechanism and the relationship system are palpable, familiar, and sometimes dramatic. I love Juliet but Juliet is a Capulet. I hate Jews but some of my best friends are Jews. Boys are yucky but after school I play with Andrew.

  Remember the study in which subjects judged a so-called grad student to have a friendly or unfriendly personality after having a brief discussion with her? The subjects made the judgment on the basis of how the young woman had behaved during the discussion, even if they knew that she had been instructed to behave that way.20 I said it made sense for the subjects to do that because they had nothing else to go on. But they did have something else to go on: they knew her (supposed) occupation, her gender, her approximate age, and her apparent race. The reason they didn’t use that categorical information to judge her personality is that she was no longer a stranger. She was someone they had talked to face-to-face, someone they now knew as an individual, someone who had been assigned a page in their people-information lexicon. To judge her personality they used the other information on the page, not the information about her occupation, age, gender, or race.

  Memories of how a given individual behaved on a given occasion are explicit—readily recalled a day later, readily put into words and passed on in the form of gossip. The relationship system tends to take up an inordinately large share of the conscious mind. Though the categorization mechanism does most of its work silently, it is more robust, more impervious to damage. An amnesiac who has forgotten having met you yesterday may still know how to use words like fish, chair, man, woman, girl, and boy.

  According to the social psychologist Susan Fiske, “Culture consists mostly of practices, skills, and motives whose cognitive representation is primarily procedural, not explicit semantic knowledge.”21 In other words, socialization consists largely of implicit, rather than explicit, learning. An important aspect of this learning is storing up information about the behavior of the members of various social categories.

  Then the stored-up information is averaged. The categorization mechanism creates prototypes by looking for central tendencies and ignoring what statisticians call “outliers.”22 The socialization system makes use of these prototypes. Children become socialized, I propose, not by imitating specific people in their lives, but by tailoring their behavior to that of the appropriate prototype.

  The ability to collect data and calculate central tendencies is not unique to our species; even birds can do it. Researchers observed collared flycatchers peering into the nests of other birds of the same species and wondered why. Turned out the birds were collecting “public information” on how the other collared flycatchers in the neighborhood were doing in terms of reproductive success. They use the information in selecting a nesting site for the following year. The researchers discovered that the birds take into account both the number and the condition of the fledglings in nearby nests. Birds “know what’s going on in their own area,” one researcher explained. Knowing what’s going on involves, in this case, collecting data by observing a number of different nests and averaging the data.23

  The birds collected public information and averaged it because it was in their best interests to do so. If they considered only a single nest—if their choice was based solely on their own reproductive success—they might make the wrong decision. Perhaps they were lucky this year. But if most of the other birds in the neighborhood also did well, that information is far more likely to be a reliable indicator of the quality of the nesting site. The more data one collects, the more accurate and useful the results are likely to be, if one has the ability to calculate central tendencies.

  The child’s situation is similar to the bird’s. Taking a single individual as her model is risky; what if she inadvertently chooses an atypical example, an outlier? It is far safer to collect as much information as her environment provides, average the data, and use the prototype as her model. I’m not saying that this is what children always do; there may be times when they imitate specific people. But that is not how they become socialized. The way children become socialized is by constructing prototypes and using the prototypes as models. The cognitive processes involved—category formation, averaging, and the acquisition of procedural knowledge—are largely implicit. These things could all be done by someone like Frederick, someone with no ability to form new episodic memories.

  This theory explains why socialization affects personality and social behavior on a society-wide (or neighborhood-wide) level. It explains why children who are reared in atypical homes may nonetheless become acceptable members of their society. It explains why siblings are no more alike in personality (once genetic similarities are taken into account) than pairs reared in separate homes. The child may have an older sibling of the same sex, and may associate with that sibling outside the home, but the sibling provides only a single data point, which is washed out by the much larger amount of data provided by people outside the family.24

  However, all members of a social category might not contribute equally to the data. Higher-status members of a
group—those who rank higher in the attention structure—are looked at more, which means they have more opportunities to contribute data. The effect would be to shift the prototype in the direction of the higher-status members.25 Rather than being the mathematical average of the group, the prototype might be a little above average.

  I’ve left out a step. Before children can tailor their behavior to the appropriate prototype, they have to figure out which one is appropriate. They therefore need to know, not only which social categories exist in their society and how the prototypical member of each category behaves, but also which category they themselves belong in. Now we come to the parts of the socialization system that are uniquely human—the bells and whistles I promised you.

  The problem that humans have to face is that a given individual can be categorized in different ways, and the categorization will depend on social context—on the situation. But let me begin with a simple case: a female child who categorizes herself as a girl. This “self-categorization,” as the Australian social psychologist John Turner calls it, causes her to favor her own social category over others and to use this social category to set the standards for her own behavior.26 What she uses as her model, according to my theory, is a prototype based on her processing of public information about her own social category.

  Like many other mental mechanisms, the socialization system provides not only the abilities but also the motivation to use them. Thus, socialization is self-socialization, not something imposed on the child by an outside force. Not, for instance, due to “peer pressure.”27 Rewards or punishments from peers or parents are not ordinarily required, because children want to be like others of their age and gender. Hence the plea, familiar to all parents, “But all the other kids are doing it.” Parents have a standard response: “If all the other kids threw themselves off a cliff, would you do it too?” The question generally receives, by way of a reply, a deep sigh of exasperation. But the parent has a point. Bison in herds sometimes do throw themselves off cliffs (though lemmings, I understand, do not).

  Children differ from bison not so much in their motivation to do what the herd does but in the complexity and flexibility of their concept of the herd. A female child does not invariably categorize herself as a girl. Depending on circumstances, she might identify herself as a fourth grader, one of the subset of good or poor students in her class, a black or white or Mexican or Chinese child, a member of a sports team or chorus, a member of her family, or simply as a child. Her behavior will be adjusted accordingly. On the school playground, where girls and boys form separate groups, she will classify herself as a girl and act girlish. In her neighborhood, where there are fewer children and boys and girls play together, she will classify herself as a child and the girlishness will be greatly reduced.28

  Anthropologists who have studied hunter-gatherer societies report similar observations. The surviving hunter-gatherer groups tend to be small and there usually aren’t enough children to divide up by sex, so girls and boys play together and there are few sex differences in behavior. But in nearby agricultural societies, where there are enough girls and boys to form separate play groups, the differences are very noticeable. Sex differences in behavior are partly biological—the same differences show up all over the world—but the inborn differences are minimized when girls and boys form a single group. When they split up into separate groups, the inborn differences are exaggerated by group contrast effects (the “anti-meme” effects I described in chapter 6). These behavioral differences have proven to be resistant both to the efforts of contemporary American parents to treat their sons and daughters alike and to the reduction of gender differences in the adult society.29

  When a person switches from one self-categorization to another, the prototype to which she tailors her behavior will also change. This, then, is one of the reasons why people behave differently in different social contexts. It also explains some of the changes that occur as children grow up. When their self-categorization shifts from “child” to “teenager,” their behavior is modified accordingly, because a different prototype has become their standard. Later, when they marry, get a job, or have a baby, their self-categorization changes again, to “grownup.” One of the reasons why people become more like their parents as they get older—especially when they become parents themselves—is that now they are members of the same social category.30

  Self-categorizations are exquisitely sensitive to social context and can change at the drop of a hat. Girls and boys in a school lunchroom or playground ordinarily categorize themselves as girls and boys, but the presence of a mean or bossy teacher can cause them to unite in a common cause and to classify themselves simply as children.31 After the terrorist attacks of September 11, 2001, Democrats and Republicans classified themselves, for a while, simply as Americans.

  As that last example illustrates, humans can identify even with groups (such as Democrats and Republicans) composed chiefly of strangers. They can identify with groups even if they don’t know who is in them. The social psychologist Henri Tajfel told some boys, supposedly on the basis of a test, that they were “overestimators” and others that they were “underestimators.” That’s all it took to evoke what Tajfel called “groupness” in the boys. When they were given the opportunity to award monetary payments to other members of the overestimator and underestimator groups (identified by group but not by name), they not only awarded more to the members of their own group: they also made sure to underpay the members of the other group.32 The boys who participated in this study all went to the same school, but none of them knew which of their classmates were overestimators and which were underestimators. There was no opportunity for the relationship system to put in its two cents. Social categorization operates independently of the relationship system, just as the system for generating the past tense of a regular verb operates independently of the system for retrieving the past tense of an irregular verb.

  Though a word like “overestimators” can create a brand new social category, categories don’t require names. Babies, as I said, show awareness of social categories before they’ve had time to learn any words. Chimpanzees, too, are aware of social categories—at least the two basic ones, “us” and “them.” Jane Goodall described what happened when the chimpanzee troop she was watching split up into two groups of unequal size. The more numerous group went to war against the smaller one, systematically picking off its members one by one and fatally injuring them, until the smaller group was wiped out.33 The fact that the members of these groups were familiar to one another—they had played together in their youth—was not enough to avert the outcome. When push came to shove, groupness overrode the relationship system.

  For a chimpanzee, familiarity is not sufficient to evoke groupness but it is necessary. A chimpanzee can never regard a stranger as one of us: an unfamiliar chimpanzee is automatically regarded as one of them. Only humans are capable of expanding the social category with which they identify beyond the circle of their close acquaintances, even beyond the circle of their tribe. The biblical story of the Good Samaritan illustrates what can happen when a human being categorizes himself simply as a human being.

  What’s unique about human groups is that they no longer have to be real groups: they can be virtual groups. The group has become a concept, an abstract category like chair or verb. The members of a social category do not have to gather together in one place; a little girl can identify herself as a girl and form a mental prototype of a girl even though she may never have seen an actual group of girls (girls usually play in pairs). Nor is personal contact required. A child can identify with a social category even if its members reject her. She can identify with a social category even if nobody but herself believes she belongs in it. Occasionally, despite the best efforts of their parents, children identify with a gender category that doesn’t correspond to their biological sex, or doesn’t correspond to the gender category to which they’ve been assigned by well-meaning but misguided physicians.34
>
  Brain size ballooned during hominid evolution. The result was a primate, Homo sapiens, that could form a concept of a group and apply the motivations and emotions originally associated with actual groups to the conceptual group. These new cognitive abilities enabled humans to form larger groups without a loss of groupness. The size of the group was now, in effect, unlimited, because its members didn’t have to be personally acquainted with one another.

  Natural selection acts on individuals, not on groups. But, as I said in the previous chapter when I was talking about schools of fish, evolution can produce patterns on the group level by tuning the characteristics of the individuals that make up the group. The fitness of a group-adapted individual depends not only on how it does relative to other individuals within its group, but also on how its group does relative to other groups. In species that go in for group-against-group warfare, such as chimpanzees, humans, and ants, bigger groups are capable of wiping out smaller ones. Hence individuals adapted to living in larger groups have an evolutionary edge. As E. O. Wilson observed, the ability to form larger groups can give individuals who inherit this ability a “decisive advantage” over their competitors.35 This decisive advantage may be the reason why we are still here and all the other hominid species—our family tree produced a fair number of them—are gone.

  “I’m not that good at relationships,” confessed the private investigator Kinsey Millhone in “N” Is for Noose.36 She was explaining why her marriages never work out and why she doesn’t want to marry, or even live with, her current lover. As I mentioned at the beginning of the book, Kinsey and I differ in this respect: I have no problem with relationships. I am on good terms with all the members of my family and have a number of friends who are dear to me. But as a child I flunked socialization.